In October 2012, a Boulder County jury convicted Michael Clark of the 1994 murder of Marty Grisham, and a judge sentenced him to life without parole. The case turned on a jar of Carmex lip balm found outside Grisham's apartment the morning after the shooting. Yvonne "Missy" Woods, a Colorado Bureau of Investigation forensic scientist eighteen years into testing evidence for the state, reported a partial DNA match to Clark and testified to it at trial. After her misconduct surfaced, prosecutors sent that evidence to an independent lab in Virginia, which could not reproduce her result. On April 11, 2025, a judge vacated the conviction — on the DNA, and on separate findings of juror misconduct and ineffective assistance of counsel. Clark walked out after twelve years in prison.
Her method left an almost bureaucratic trail. Investigators found she had lowered the "quantification value" — the number telling an analyst how much DNA a sample contains — to make it appear that no DNA was present, or too little to be worth amplifying, and then moved the case along. In thirty-five sexual-assault cases and at least seven homicides, her reports said no male DNA was found when male DNA had been detected or contamination was possible. Asked why, she said it was a rush batch and she was trying to get data out. She did not invent matches or fabricate profiles; the bureau's own review was explicit about that. She deleted the parts of the record that would have required her to keep working, and that was enough to put more than a thousand cases under review — assaults, rapes, murders, all trusted because a state lab had certified them.
What makes this harder than an ordinary fraud story is that Colorado had the truth years before it acted on it. In 2014, a coworker questioned her testing in a case and reported it to a technical leader. Nothing changed. In 2018, she was accused of manipulating data, pulled off casework while the accusation was reviewed, then reinstated — and the findings of that review never reached the bureau's director or the state public-safety leadership above him. CBI has since said it is investigating why not. It took an undergraduate intern, who noticed an anomaly in her data during a research project in September 2023, to force the case open for good.
Woods pleaded guilty in June 2026 to four felonies — cybercrime, perjury, forgery, and attempting to influence a public servant — under a deal that guarantees prison rather than probation. She is scheduled to be sentenced on September 8, 2026, facing eight to sixteen years. "Today, Ms. Woods accepted responsibility not only for individual acts of misconduct but for the full scope of criminal conduct that spanned decades," said Alexis King, the district attorney who prosecuted her. Colorado has spent millions of dollars retesting the backlog her shortcuts created, and Clark's own case is still unresolved: prosecutors said in September 2025 they would try him again, and in May 2026 his lawyers moved to dismiss, arguing three decades of delay had left no fair trial possible.
A crime lab exists to answer one question honestly: what does the evidence actually show? Every institution that certifies something on the public's behalf — a bank's auditor, a building's inspector, a drug's regulator — survives on that promise, and fails it the same way: not because nobody noticed the lie, but because whoever noticed wasn't believed, or wasn't followed up on, until ignoring the problem cost more than confronting it. The first report that something was wrong with her work reached a supervisor in 2014. Michael Clark had been in prison two years by then. He stayed for eleven more.